Variations of Two-Transistor Circuits: A Tribute to the Versatility of MOSFETs | Hacker News Reader

Variations of Two-Transistor Circuits: A Tribute to the Versatility of MOSFETs

researchgate.net

82 pointsjrwan34 comments

Some of these are standard, a few are genuinely clever, but many are bad designs and I'd never let someone tape them out in production.

2. Connecting the body terminal to something other than the highest voltage is dangerous and we rarely do it except in switches and diff pairs. We'd never do something like this. You can't even connect the nmos body to the gate unless you're using "deep" nwells which is an expensive process step that most people seem to avoid these days

3. I hate it when foundries do this in their standard cell libraries. Its an extremely weak pulldown and I've seen problems in production caused by using this structure. You can make a much better structure with a third transistor and positive feedback where devices still aren't connected to the supplies, but you get a stronger pull.

4-6. These are current-mode logic, but you never see them made this way because they're slow and high power. Instead they're usually made with a tail current source on the "bottom" and resistors on top, which keeps all transistors in a faster operating region. They get used often in RF clock dividers.

16. Again, connecting the body is dangerous, especially when the potential is somewhat unknown, but this does get used in extremely low power/low voltage applications.

I didn't get through all of them, but it made me wonder - with two 4-terminal devices, how many possible configurations can you actually make?

Which are clever?

18 is cool. Folding is a really under-rated and under-utilized technique. Common-gate impedance matching is "standard", but 22 uses it to also generate differential current outputs which is cool - I imagine it sees more use in RF design. 23 is also interesting. 27 blew my mind. I'm not familiar with ultra low power techniques so that was cool to see.

The paper cites https://ieeexplore.ieee.org/abstract/document/8506644, which enumerates 582 possible topologies for two transistors, and 56280 for three. I imagine most are useless and/or dangerous.

> a few are genuinely clever

Another perspective is that it would take a computer only a few minutes to generate all these circuits and test them for usefulness in a simulator.

That's harder than you might think. How would you test for all possible uses? Even if you looked at something fundamental like an N-port, you'd still have to define metrics for usefulness and many of the topologies in the paper would evade most obvious circuit metrics.

Well, you are right to some extent, but in practice you don't need "all possible uses" and instead you are looking for a specific behavior, which you can easily encode in a test that uses a simulator.

The article mentions that circuit 23 was found using an "exhaustive search", so presumably using an automated means.

But I don't think that works in the general case to find all "useful" circuits. For example figure 27 is a pA source and requires 2 precise control voltages and a clock source. 39 is a peak detector. Does not seem that easily encoded.

If you happen to play with electronics, most circuits in the paper cannot be built at home realistically. Those circuits work best with MOSFETs built in silicon, with very precise tolerances.

For example the first circuit; the inverter; when Vin is low, the bottom NMOS is off, the top PMOS is on. And vice versa when Vin is at the high voltage suitable for the devices.

Statically this works pretty well. But the trouble appear during the transition of Vin between the two state.

Un the middle of the range both MOSFETs are likely conducting, effectively producing a dead short!

On silicon they will design both MOSFETs to control and reduce this effect as much as possible. If you attempt to reproduce this, you might be in for a magic smoke release moment.

I did manage to somewhat reproduce this without a constant meltdown with tiny discrete MOSFETs by using two tiny ones with the most complementary specifications as possible. The goal was to have them with the least amount of overlap in conduction with regard to their respective inputs.

The same thing would happen on silicon. It's just a question of transition time. I've seen power mosfet inverters on some board designs, but you're probably better off using a discrete logic inverter because it would be cheaper...

Can you use non-complimentary MOSFETS if they have a dead zone (i.e. both are closed for some middle voltage, resultig in high-z output)?

Yes, I was alluding to it in my post. I did successfully build a MOSFET driver (to drive a big MOSFET for a DC-DC converter) using two tiny MOSFETs (tiny as in sot23 package) in push-pull configuration. I found two MOSFETs that did barely overlap. It seems to work well. And the power dissipation was very reasonable at 42kHz. Note that I am merely a hobbyist.

Using non well matched MOSFETs resulted in a rapid uncontrolled atomization of the MOSFETs.

> Un the middle of the range both MOSFETs are likely conducting, effectively producing a dead short!

This is exactly the configuration described in the CD4069UBE datasheet, where single stages of the hex inverter are biased to the midpoint with a feedback resistor. This turns them into a surprisingly good programmable inverting amplifier, rather like an opamp without a non-inverting input.

You can buy push-pull configured transistor pairs in a package.

CD4007 comes to my mind, but that one is ancient (still available though in a breadboard friendly package).

What layperson-friendly tools are out there to simulate simple analog circuits? For years now I've been going to the ancient app at https://www.falstad.com/circuit/ but wondering that surely there must be something that Real Electronics Builders use. (I intentionally did not say Engineers because I assume that would be covered by the pricy and/or UI-challenged tools like spice and company).

I'm afraid there really isn't any hobbiest simulators out there. A lot of these arrangements rely on the intrinsic characteristics and geometry of the MOSFETs used.

For simple circuits, LTspice: https://www.analog.com/en/design-center/design-tools-and-cal...

And it's definitely used by real engineers. It's for making switching power supplies, so it can handle fairly complex circuits. I wouldn't venture into RF/high speed circuits with it though, since the included parasitics aren't sufficient.

And, custom functions are supported, so, I believe, you can put whatever maths in a component (I've never ventured this far): https://ltwiki.org/index.php?title=B_sources_(complete_refer...

The parasitics all depend on the model, as I used LTSpice recently for an RFIC class for a 5 GHz LNA. The .net directive allows two-port analysis and even a Smith chart. Though Microwave Office is my go-to tool.

One can also use KiCAD to generate netlists for other SPICE tools like ngspice.

Most EDA packages will include some sort of simulator, often SPICE-based. The fancier ones for RF work include or integrate with field solvers.

circuitlab.com is great for simple designs. Not as powerful as some SPICE simulators, but amazingly convenient.

(CircuitLab developer here.) Thank you for the kind words! :) Clickable link: https://www.circuitlab.com/ if anyone wants to give it a try.

As far as the linked article "Fifty Nifty Variations of Two-Transistor Circuits: A tribute to the versatility of MOSFETs", I will mention that CircuitLab does NOT currently provide a MOSFET model that includes a separate body terminal.

Instead, the body (also sometimes called the "back gate") and source terminals are assumed to always be internally connected in the CircuitLab MOSFET models, resulting in a three-terminal device.

Almost all real-world discrete MOSFETs you can buy are also three-terminal (gate/drain/source), not four (gate/drain/source/body). Some discussion here https://electronics.stackexchange.com/questions/137161/why-a... and here https://electronics.stackexchange.com/questions/185109/mosfe... but not especially satisfying in my opinion.

As a result, Figures 2, 16, 27, 29, 46, 49, and 50 can NOT currently be realized within CircuitLab.

There is JADE (for education), which can be integrated into a blog or for teaching on the web. It supports hierarchical custom analog and digital modules, a convenient way to create test vectors, (docs for that are floating around) This tool was used by students to build and grade the BETA CPU.

https://computationstructures.org/exercises/sandboxes/jade.h...

I like how I've been able to use MOSFETs as switches for the past 5 years while being almost completley unaware of the other things it can do. To me, it's a component that uses a small (low-voltage, low-current, low-resistance) signal to switch on a much larger path. Anything that makes it not behave like that idea is just problems for me (I spent a better part of a day debugging a circuit that works fine with a 5V arduino but not a 3.3V ESP32. I don't think I've even built a circuit with two transistors (multistage amp, flipflops...)

Discrete power MOSFETs are only switches. Discrete linear-stable MOSFETs are also usable as linear amplifiers. TFA is about mosfets on ICs. When making your own ICs, you have precise control over a lot of the variables, and the variables you have less precise control over are often going to be very similar for adjacent FETs on the same wafer, so you can do things with them that you can't do with discrete MOSFETs.

What stinks is that linear use has waned so much that most lateral mosfets are no longer produced. Everything is vertical switching for PWM control.

Wow, you aren't kidding. I just checked Onsemi and Vishay and neither makes lateral mosfets anymore. I'm surprised because many audio amplifiers still have a linear output stage, are those just always BJTs now?

Most of the linear amps these days are integrated devices that need a few supporting components and you feed a line level audio signal. Though you can still get decently sized BJT's and you can build your own amps.

I could have sworn I'd seen a reference to someone working out or cataloging all known-to-be-useful two-transistor circuits with bipolar transistors but I can't seem to find it now. Anyone know?

This is a fun reminder that a MOSFET is useful as a transistor, a diode, and a capacitor.

And sometimes, as a fuse.

And resistor. And if etched in spiral the should've inductance, so also coils.

Also various types of sensors (magnetic, chemical, radiation...)

I'm getting a CAPTCHA page forever. Is there an open-access copy of this paper?

Reply on news.ycombinator.com